1. X-linked retinoschisis (XLRS) gene therapy pre-clinical study and clinical trials XLRS is a genetic disease caused by mutations in the retinoschisin (RS1) gene. An absent or mutated RS1 protein in the retina of patients with XLRS leads to abnormalities in the normal laminar structure of the retina, resulting in impaired visual acuity and increased propensity to retinal detachment. Gene replacement therapy holds the promise of curing the disease. In collaboration with Dr. Paul Sieving's group, we have developed and optimized adeno-associated viral (AAV) vectors that are capable of mediating stable and retinal specific expression of human RS1 protein in mice. To meet the FDA requirements for gene therapy clinical trials, Dr. Sieving's group and us have completed a pre-clinical efficacy study in a retinoschisin knockout mouse model, and a vector toxicology study in both mouse and rabbit models. Investigational New Drug (IND) application to FDA has been approved and a phase I/II clinical trials are ongoing. As of August 2019, a total of twelve XLRS patients have received the vector administration. The initial clinical findings in treated patients have recently been published in Molecular Therapy. We are now working on improving the efficiency of gene delivery for better treatment effects in patients. 2. Pre-clinical gene therapy studies for retinitis pigmentosa due to RPGR or RP2 mutations X-linked forms of retinitis pigmentosa (XLRP) are relatively severe blinding disorders, resulting from progressive photoreceptor dysfunction primarily caused by mutations in RPGR or RP2 gene. Gene therapy for RPGR-XLRP was hampered by the relatively slow disease progression in mouse models and by difficulties in cloning the RPGR-ORF15 cDNA that includes a purine-rich 3-coding region. We managed to overcome these problems and have generated AAV vectors carrying full-length mouse and human RPGR ORF15 coding sequences. We have also developed a self-complementary AAV vector carrying human RP2 expression cassette. We have completed proof-of-principle long-term (18-24 months) dose efficacy/toxicity studies in mouse models with RPGR or RP2 deficiency. Our results demonstrate that administration of the RPGR AAV vectors at appropriate doses can significantly preserve the retinal function and delay the photoreceptor loss in the mice with RPGR deficiency. Additionally, administration of the RP2 vector with a broad dose range can remarkable maintain the function and viability of cone photoreceptors in the mice with RP2 deficiency. These results lead to two papers published in Human Molecular Genetics. In order to bring the therapies to the clinic, we have done the following during the 2019 fiscal year. 1) Searching for early disease hallmarks in Rpgr-mutant mouse line and other lines with retinal degeneration. We have identified an outer retina change in OCT as a common and early hallmark in 6 mouse lines with retinal degeneration, and gene replacement therapy could reverse this OCT change. This new finding has been submitted for publication. 2) Conducting dose-efficacy in a mouse Rpgr mutant model with a faster retinal degeneration.The mouse model we used has an Rpgr-null background with one allele of Cep290 rd16 mutation, and this work is nearly completed. 3) Comparing two AAV serotypes in non-human primates (NHPs) for rod and cone photoreceptor transduction. This work is a collaboration between NEI and Charles River. The animal study has been completed and the data analysis is ongoing. Our initial analysis show that both AAV8 and AAV9 were able to transduce cone photoreceptors efficiently in NHPs at the vector doses we tested. 4) Investigating disease progression in heterozygous female Rp2 KO mice and testing whether gene replacement is efficacious for treatment in these mice. This work is ongoing. 3. Gene therapy for Leber congenital amaurosis due to CEP290 mutations Leber congenital amaurosis (LCA) is one of the most common causes of blindness in children. People with this disease typically have severe visual impairment beginning in infancy. Mutations in the CEP290 gene account for 20-25 percent of LCA, afflicting an estimated 20,000 people worldwide. Since the size of CEP290 coding sequence (7.4 kb) exceeds the packaging limit of AAV vector (5kb), we have been seeking alternative approaches for treating CEP290-LCA. In particular, we tested whether delivery of truncated functional domains of Cep290 could complement certain types of CEP290 mutations. We have made a series of AAV vectors carrying different portions of Cep290 coding sequence and have tested them in a mouse model with Cep290 mutation. We recently identified one CEP290 fragment that is capable of preserving the retinal function and structure in the mouse model with a c-terminal in-frame deletion (rd16). This fragment, together with the hypomorphic form of the mutant CEP290, reconstitute full function of CEP290 in a cell/tissue independent manner. The manuscript of these findings has been published in Cell Reports during this fiscal year. In addition, we are collaborating with labs in NCATS on small molecule and siRNA screen, in order to find better ways to delivery large genes by AAV vectors. 4. In vivo applications of CRISPR/Cas9 for treating retinal degenerative diseases CRISPR/Cas9 mediated genome editing has been rapidly advancing in recent years. However, its applications in postmitotic photoreceptors-the direct target of a majority of acquired or inherited retinal degeneration, has been limited. We have recently established an AAV-delivered photoreceptor-specific CRISPR/Cas9 system. The system is highly efficient in modifying photoreceptor genome, as exemplified by disrupting EGFP or Nrl gene following vector administration to mice at postnatal day 14. These results have been published in Nature Communications. During this fiscal year, we have been using the AAV-CRISPR system to develop treatment for an autosomal dominant retinal diseases--retinitis pigmentosa caused by rhodopsin mutations. We tested if the CRISPR/Cas9 approach could specifically disrupt the T17M human allele in a transgenic mouse line and rescue the disease phenotype. Three Cas9 nucleases, spCas9, espCas9 and HifiCas9 were used and compared. The Cas9 and the sgRNA expression cassettes were packaged into AAV8 respectively. Mice received subretinal administration of vectors. We found that the SpCas9 vector together with the sgRNA-T17M vector resulted in markedly higher ERG amplitudes in the therapeutic eyes than in the control eyes in the T17M mice. However, sanger sequencing revealed both on-target and off-target events. Treatment of the eSpCas9 vector together with the sgRNA-T17M resulted in low on-target activity while no off-target events were observed. The treatment of HifiCas9 vector together with the sgRNA-T17M demonstrated superior on-target activity and high specificity. Our current work is focused on collecting data using the HifiCas9. 5. Core function: AAV vector design, production and administration for collaborating groups During the 2019 Fiscal Year, we have produced 28 AAV vector preparations to support the research projects of our own and over ten collaborating groups including NIH intramural, extramural and international research groups.